Create ESGmodel
Browse files
ESGmodel
ADDED
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
|
2 |
+
|
3 |
+
|
4 |
+
|
5 |
+
import torch
|
6 |
+
from transformers import AutoTokenizer, AutoModelForSequenceClassification
|
7 |
+
|
8 |
+
# 加载第一个模型
|
9 |
+
tokenizer1 = AutoTokenizer.from_pretrained("Emma0123/fine_tuned_model")
|
10 |
+
model1 = AutoModelForSequenceClassification.from_pretrained("Emma0123/fine_tuned_model")
|
11 |
+
|
12 |
+
# 加载第二个模型
|
13 |
+
tokenizer2 = AutoTokenizer.from_pretrained("jonas/roberta-base-finetuned-sdg")
|
14 |
+
model2 = AutoModelForSequenceClassification.from_pretrained("jonas/roberta-base-finetuned-sdg")
|
15 |
+
|
16 |
+
# 输入文本
|
17 |
+
input_text = input()
|
18 |
+
|
19 |
+
# 对第一个模型进行推理
|
20 |
+
inputs = tokenizer1(input_text, return_tensors="pt", truncation=True)
|
21 |
+
outputs = model1(**inputs)
|
22 |
+
predictions = torch.argmax(outputs.logits, dim=1).item()
|
23 |
+
|
24 |
+
# 根据第一个模型的输出进行条件判断
|
25 |
+
if predictions == 1:
|
26 |
+
# 使用第二个模型进行判断
|
27 |
+
inputs2 = tokenizer2(input_text, return_tensors="pt", truncation=True)
|
28 |
+
outputs2 = model2(**inputs2)
|
29 |
+
predictions2 = torch.argmax(outputs2.logits, dim=1).item()
|
30 |
+
print("Second model prediction:", predictions2)
|
31 |
+
else:
|
32 |
+
print("This content is unrelated to Environment.")
|